Model of Customer Loss Based on Cluster and Neural Network

Article Preview

Abstract:

In telecom CRM, characters of customer loss are uncertain and complicated. It is difficult to describe using a single algorithm. Both Two-Step clustering algorithms and neural networks analyzing method are adopted in this paper, and build a Telecommunications customer loss prediction model. When it is used in actual area, the rate of accuracy is up to 80.12%.

You might also be interested in these eBooks

Info:

Periodical:

Pages:

284-287

Citation:

Online since:

July 2012

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2012 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] Xu YuanChun, Cheng ZhaoHan, Liu BingXiang. A analysis method of customer losing crisis based on the decision tree. The computer and modernization [J], 2004,8:1-4.

Google Scholar

[2] Cao YaDong, Mo LiPing. The customers keeping -a new topic in telecom enterprise [N], People's post newspaper, 2003.

Google Scholar

[3] Xue Wei, The Clementine data mining method and application, publishing house of electronics industry, 2010-9.

Google Scholar

[4] Xiong Ping , Data mining algorithm and the Clementine practice, Tsinghua university press,2011-4.

Google Scholar

[5] Ji XiYu, Application example of data mining technology, mechanical industry press, 2009-4.

Google Scholar

[6] Xie BangChang, Clementine data mining application practice, mechanical industry press,2008-3.

Google Scholar

[7] Wang Ying , Chen ZhiPing, The application of the losing telecom customers combined with K-means classification method ,Jiamusi university journal (natural science edition) , 2010.

Google Scholar